dataset_info:
features:
- name: question
dtype: string - name: gold
dtype: string - name: target
dtype: string - name: prediction
dtype: string - name: subset
dtype: string - name: qwen_extracted_answer
dtype: string - name: qwen_score
dtype: float64 - name: harness_extracted_answer
dtype: string - name: harness_score
dtype: float64 - name: lighteval-c24870ea_extracted_answer
dtype: string - name: lighteval-c24870ea_score
dtype: float64
splits: - name: train
num_bytes: 3413283
num_examples: 1324
download_size: 1345415
dataset_size: 3413283
configs: - config_name: default
data_files:- split: train
path: data/train-*
- split: train